Russia (bbabo.net), - Russian scientists have attracted neural networks to save the forests of our planet. With the help of artificial intelligence, a team of specialists, which included representatives of the Siberian Federal University, learned to identify trees affected by pests already in the early stages of infection. The team's study was published in the Q1 "IEEE Access" journal.

In the fight against global warming, fires and other factors that threaten the Earth's forests with mass extinction, scientists this time focused on dangerous pests. The study, according to the team, will make it possible to take timely preventive measures in case of infection of trees.

To diagnose the forest, experts used unmanned aerial vehicles and YOLO (You Only Look Once) deep neural network algorithms, which allow you to quickly identify objects in real time. These objects were, of course, trees - in this case, Norwegian or European spruce, common in many countries.

Scientists compared different architectures of neural networks, tested them and determined the best option for detecting infected trees, said Anastasia Safonova, Associate Professor of the Department of Artificial Intelligence Systems of the Siberian Federal University, Candidate of Technical Sciences.

The university clarified that artificial intelligence can be configured to detect various pests and any type of tree. This will allow taking measures to prevent "forest epidemics".

bbabo.Net